Online Program Home
My Program

Abstract Details

Activity Number: 359 - Contributed Poster Presentations: Biopharmacutical Section
Type: Contributed
Date/Time: Tuesday, July 31, 2018 : 10:30 AM to 12:20 PM
Sponsor: Biopharmaceutical Section
Abstract #327044
Title: Assessing Reproducibility When Making Mid-Course Changes in Clinical Trials Based on External Data
Author(s): Yingqi Shi* and Grace Gao and Keith Karcher
Companies: Johnson & Johnson-Janssen R&D and Janssen R&D and Janssen R&D
Keywords: reproducibility; posterior predictive distribution; MCMC
Abstract:

Testing early treatment interventions in clinical trials, especially for slow progression diseases, a long trial duration is required to assess efficacy on clinical outcomes. Oftentimes, due to the time constrain, it is not feasible to conduct a phase 2 efficacy study before starting a large pivotal study. Thus, due to lack of phase 2 data, Knowledge on the key design parameters for the pivotal study is limited. If relevant external data will be emerging during the ongoing pivotal study, then the key design parameters can be updated based on the emerging external data while still maintain the study blind. However, adapting based on external data could be risky. The reproducibility of the results of the external data is not guaranteed. Sometime the external data may not quite resemble the study population. There is a need to assess whether the results from the external data can be reasonably reproduced in the ongoing study. One way to assess the reproducibility is to evaluate the samples from the posterior predictive distribution given the external data. This can be easily accomplished via MCMC. Examples and simulations are provided to evaluate the utility of such an approach.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2018 program